Autonomous sound monitoring shows higher use of Amazon old growth than secondary forest by parrots

Abstract

Forest regeneration may reduce the current loss of species due to tropical deforestation, but little is known about the extent and inter-specific variability of this effect. We compared the probability with which nine parrot species use old-growth and secondary forests in a ~400km2 Amazonian landscape, while considering two types of habitat use: perching and flyover use. Perching use, when individuals stop at a sampling site, implies resting, foraging, or breeding activities; flyover merely implies that parrots fly through, above the canopy at a site. We established 155 sampling sites spanning old growth and approximately 30-year-old secondary forest, and sampled them repeatedly using autonomous audio recorders. Perching and flyover use were distinguished based on the temporal variation of intensity in parrot vocalization sonograms. We modeled our data with a set of species-specific, multi-state occupancy models that estimate the probability of each type of use for both habitats while accounting for imperfect species detection. Models were fit in a maximum likelihood framework and ranked according to the Akaike information criterion. All but one species fly over both habitats with the same probability, while seven out of nine show a higher probability of perching in old growth than in secondary forest. Interspecific variation in response to habitat change was not explained by variation in body mass or relative brain size. After three decades of forest regeneration in our study area, there are still measurable differences in habitat use, with a broad tendency for parrots to favor old growth over secondary forest. © 2015 Elsevier Ltd.